This effect still occurred when I was operating at f/10 and higher on a tightly-collimated SCT - forgetting about lenses altogether.
I don't recall having a purple ring aspect to my D70 years ago. Before or after I modified it.
I saw a slight amount of magenta stars in a few of my recent night shots but then I see similar or worse using my Pentax 67 lenses with my monochrome Proline 16803.
I take that as a bit of CA. I don't see how the sensor causes purple fringing. There is no evidence of purple fringing in day shots.
The 800E in particular shows up CA in lenses that in the past were considered top notch. Its the "downside" of such powerful resolution.
It corrects quite easily using Photoshop selective colour adjusting the magenta and cyan sliders. I could even record an action to suit so I could run it on any shot I did with the camera.
If you look at Waynes images the 14-24 does not show much fringing at all (its the better of the lenses) and the Samyang shows more. I don't how this lens rates but it would appear to have more CA.
I could do a comparo as I have 40D and modded 20D as well with a Novoflex adapter so I could use the same lens.
It may take a little bit to do one but seeing as there is interest I'll make the time to do one.
Greg.
Last edited by gregbradley; 14-05-2012 at 03:36 PM.
You also get purple fringing in the brighter stars with Canon 5D2 + certain lens and aperture combos. It's pretty evident with the Canon 24 1.4 II and 85 1.2 below f/2.8, less so but still there at f/2.8. With the Nikon 14-24 + adaptor @ f/2.8 it's pretty much gone.
I normally fix it by reducing the 'purple' saturation in ACR.
With the better quality ISO on these new cameras (inc. 5DIII) it's much more possible to stop lenses down and still obtain great low light images now. For every criteria that I look at, these cameras have such great direct and flow-on improvements, such as this example.
I think you are right, I have seen images with Samyang 35mm on Canon body (5DMKII) and it is full of chromatic aberrations at wide apertures as well. A good test would be to put the Samyang on your D800 and test at apertures from 1.4 to 8 doubling the ISO as you close the aperture by one stop.
If you look at Waynes images the 14-24 does not show much fringing at all (its the better of the lenses) and the Samyang shows more. I don't how this lens rates but it would appear to have more CA.
I could do a comparo as I have 40D and modded 20D as well with a Novoflex adapter so I could use the same lens.
It may take a little bit to do one but seeing as there is interest I'll make the time to do one.
Greg.
I don't notice much of the CA with 14-24 at f/2.8 either. That would be great if you could run tests with Samyang 35mm and Nikkor 14-24 on Canon bodies and then on Nikon D800E in DX mode (to make the playing field level). It would highlightr any differences in Canon vs Nikon sensor/microlenses and show if modded 20D is more prone to violet fringing than normal non-modded body.
This effect still occurred when I was operating at f/10 and higher on a tightly-collimated SCT - forgetting about lenses altogether.
Chris,
Even perfectly corrected SCTs have spehochromatism induced by the corrector plate ( http://www.telescope-optics.net/SCT.htm ) and whilst I have no experience with imaging through an SCT, it could be that the UV/IR cutoff filter (hotmirror) is more agressive in certain wavelengths with some camera bodies and it "masks" the problem.
Alex, I guess that my point is this. The same imaging train, whatever the case (same Nikkor lens on both Nikon and Canon body) or the SCT - the Nikons I have always produce the purple flare - in every situation. The Canons do not (350D, 450D, 40D and 5D-II). I'm not saying that the Canon doesn't produce chromatic aberrations, of course they do, but not NEARLY to the same extent as all of my Nikon bodies.
I'm ecstatic with the camera! Wait until you see the images I took last night!
You won't be able to buy it for $3300 though - that's a US price with no Aus taxes. Here they are $3999 full price, but I got mine a bit cheaper and so did someone else (as loyal camera shop regulars!). The D800 is a few hundred cheaper of course too.
I'll try to post a dark frame for you soon.
Thanks,
Wayne
Thanks - I look forward to seeing a dark frame & how
many hot pixels & the noise you get.
Alex, I guess that my point is this. The same imaging train, whatever the case (same Nikkor lens on both Nikon and Canon body) or the SCT - the Nikons I have always produce the purple flare - in every situation. The Canons do not (350D, 450D, 40D and 5D-II). I'm not saying that the Canon doesn't produce chromatic aberrations, of course they do, but not NEARLY to the same extent as all of my Nikon bodies.
Chris,
I don't doubt it and it could be due to the difference in UV/IR cutoff wavelengths or CA induced by microlenses or something else. 36MP of D800 will show that defect more when viewed at 100%,
I had zero purple flares when D700 was used with f/3.6 reflector, no matter how strongly overexposed the bright stars were. Then again D700 is not every Nikon and we would never know for sure what is going on inside Nikon RAW files, because they are "pre-cooked" by Nikon and not purely RAW unlike Canon.
I've uploaded some images below that show that the purple flaring I have in my images is due to the lens. Luckily I had done some trials of different f-stops with the Samyang 35mm last night on the same part of the sky! Four crops at f1.4, f1.8, f2 and f2.8 are shown below via the links. The only problem is that they are not necessarily the same exposures - I hadn't planned on doing this test! - but it still shows the flaring problem. It's most evident at f1.4 and f1.8 but not so evident at f2.8 (even though I managed to move the camera slightly that elongated the stars!).
Crops from images of the Nikon 14-24 at 14mm and the Nikon 80-200 at both 80mm and 200mm below don't show any purple flaring. There is slight blue flaring in both these lenses however when used wide open at f2.8.
I've uploaded some images below that show that the purple flaring I have in my images is due to the lens.
Thanks for those, Wayne - very similar with my experience:
14-24 and 70-200 wide open at f/2.8 - no purple, some minor blue CA with both D700 and D3s. Looking at Nikkor 35mm f/1.4 lens stopped down to f/5.6 - not much fringing even in the very corner of the frame (30sec, f/5.6, 4350K WB, D3s, 100% crop attached, no lens corrections applied to the RAW file).
I think fringing depends more on the Camera/Lens/f stop or Camera/Telescope system as a whole; and variables such as the lens correction, light cone angle, UV/IR cutoff filter characteristics and microlenses are determining the outcome.
Thanks for your example that backs up mine Alex. It looks like a characteristic of the Samyang 35mm. And as you said probably due to a range of things in the system.
I certainly do but I didn't think non-D800 users could open them as each new camera RAW files are different and people need the latest versions of software, unless they have an updated Photoshop now. Let me know if that's incorrect!
I certainly do but I didn't think non-D800 users could open them as each new camera RAW files are different and people need the latest versions of software, unless they have an updated Photoshop now. Let me know if that's incorrect!
